token.app › Thinkingmachines › Inkling
Inkling pricing & benchmarks
Inkling is available from Thinkingmachines at $1.00 per million input tokens and $4.05 per million output tokens ($1.76 blended at 3:1). It accepts up to 524,288 tokens of context. Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total. It is designed for general-purpose reasoning, coding, agentic and tool-use systems,...
Pricing
Per million tokens · OpenRouter
Input$1.00
Output$4.05
Blended (3:1)$1.76
Specification
thinkingmachines/inkling
Context window524K
Max output472K
ReleasedJul 2026
LicenceOpen weights
Modalitiesin:text, in:image, in:audio, out:text
Benchmarks
Independently run — not vendor self-reports. Every score links its source.
| Benchmark | Score | Config | Run | Source |
|---|---|---|---|---|
| GPQA Diamond | 88.3% ±1.9 | xhigh effort | 2026-08-05 | Independently run · Epoch AI |
| FrontierMath T1–3 | 33.3% ±2.8 | xhigh effort | 2026-08-06 | Independently run · Epoch AI |
| FrontierMath T4 | 4.9% ±3.4 | xhigh effort | 2026-08-06 | Independently run · Epoch AI |
| OTIS Mock AIME | 88.9% ±3.9 | xhigh effort | 2026-08-05 | Independently run · Epoch AI |
Scores published by Epoch AI, 'AI Benchmarking Hub', used under CC BY 4.0.
Cheaper models that score at least as well on GPQA Diamond
Strictly lower blended price AND an equal-or-higher GPQA Diamond score. One benchmark is one dimension — a model that wins here may still be weaker at your specific task.
| Model | Provider | Blended $/1M | GPQA Diamond | |
|---|---|---|---|---|
| GPT-5.6 Luna | OpenAI | $0.450 (−74%) | 91.6% | Compare → |
| MiniMax M3 | MiniMax | $0.525 (−70%) | 90.9% | Compare → |
| Inkling Small | Thinkingmachines | $0.637 (−64%) | 88.5% | Compare → |
| DeepSeek V4 Flash 0731 | DeepSeek | $0.660 (−63%) | 91.0% | Compare → |
| Qwen3.6 Plus | Alibaba | $0.731 (−59%) | 88.4% | Compare → |
Where to run Inkling
2 hosts · all priced within ~5% of each other. Blended at 3:1. ⚠ marks an option serving less context than the best available, or re-pricing above a prompt-length threshold. Bracketed labels are the provider's own service tier or region.
| Host | Blended $/1M | In / Out | Context | Weights | Uptime 24h |
|---|---|---|---|---|---|
| DeepInfra | $1.72 | $0.950 / $4.05 | 524K | fp8 | 91.1% |
| Together | $1.76 | $1.00 / $4.05 | 524K | — | 94.8% |
Host prices and uptime from OpenRouter, synced 2026-10-22. Lower-precision weights (fp4, fp8) can cost less and score worse than the same model at bf16 — the score above is not host-specific.
Compare Inkling
FAQ
How much does Inkling cost?
Inkling costs $1.00 per million input tokens and $4.05 per million output tokens on OpenRouter. At a 3:1 input:output mix that blends to $1.76 per million tokens.
What is Inkling's context window?
Inkling accepts up to 524,288 tokens of context, and can generate up to 471,859 output tokens.
How good is Inkling on benchmarks?
Inkling scores 88.3% on GPQA Diamond, independently run and published by Epoch AI. 4 benchmark results are listed on this page, each with its run date and source.
Is there a cheaper model as good as Inkling?
Yes — 5 models in our catalogue cost less per token than Inkling and score at least as high on the same benchmark. The cheapest is GPT-5.6 Luna at $0.450 per million tokens blended.